SUMMARY (Extended): Networker 4.4 SNAFU

From: V.S. Glukhov <glukhov_at_sequence.Stanford.EDU>
Date: Wed, 15 Apr 1998 08:28:15 -0700 (PDT)

An angry subscriber told me that I forgot to include the original message.
Sorry. I also received several helpful hints regarding TZ88*
firmware upgrade.

=====================================================================
Original message (edited):

I am having this weird problem with NSR 4.4 after upgrading to
DU 4.0D. Here is the excerpt from the log file:

-----------------

 4/13/98 5:38:20 nsrd: mymachine.stanford.edu:<7>/assembly43 saving to
pool 'CaveFull' (CaveFull.009)
 4/13/98 5:48:09 nsrd: media warning: /dev/nrmt1h writing: end of
media, at file 76 record 1094
 4/13/98 5:48:09 nsrd: media notice: tz88 tape CaveFull.009 on
/dev/nrmt1h is full
 4/13/98 5:48:10 nsrd: media notice: tz88 tape CaveFull.009 used 30 GB
of 40 GB capacity
 4/13/98 5:49:25 nsrd: media warning: verification of volume
"CaveFull.009", volid 6500 failed, read open error, Device busy
 4/13/98 5:49:25 nsrd: media notice: verification of volume
"CaveFull.009", volid 6500 failed, volume is being marked as full.
 4/13/98 5:49:25 nsrd: media notice: Save set (21814)
mymachine.stanford.edu:<7>/assembly43 volume CaveFull.009 on /dev/nrmt1h
is
 being terminated because: media ve...
 
-----------------

The machine is AS 600 5/333. Firmware 5.0 applied. The device is TZ887
(Quantum DLT/Digital DLT) "jukebox", tape drive is defined as tz88. "mt
-f /dev/nrmt1h status" does not work until the busy state is cleared
(obviously). jbexercise and tapeexercise do not give any errors (after
nsr is restarted to clear the "device busy" state). nsrjb -H and nsrjb
-I do not give any error (except "Cleaning tape" is not in media
database) once nsr is restarted. No SCSI, IO or other errors reported by
uerf. The group being backed up is two (local and remote) DU 4.0D
machines with BRX440 kit installed. All other groups are disabled (for
other reasons). I am at the last stage of recovering from a pretty big
crash and trying to perform a full backup "as soon as possible".

Should I downgrade to 4.3(something)?

Thanks.

V.S.G.

SUMMARY draft #1 (edited):

>
> Thanks to all who replied, especially to Tom Webster.
>
> Tom's reply explained it all:
> ----
> When we came across this problem with our TL812 and NSR4.4, DEC's NSR
> support indicated that it was a known problem and could be fixed by
> upgrading the firmware on the tape drives in the jukebox to the latest
> version ("v100"). Unfortunately, our FSE has been unable to build a good
> update-image tape and is trying to get one shipped out to us.
> ---
>
> Well, I did find v100 firmware image somewhere deep in Quantum's ftp site.
> There's even a note on how to make a firmware upgrade tape (dd is our
> friend) and how to change the product/vendor id. I was willing to
> give it a try, but Storageworks manuals neglected to mention how to upgrade
> TZ887's firmware using FU tape.
>
> I downgraded Networker to 4.3 and everything worked fine.
>
>
>
> V.S.G.
>
>
>

TZ88 FIRMWARE HINTS:

Alan Rollow _at_ DEC:

>If you have the correct firmware image, you can created a firmware
>update tape using the ltf command. The TZ88 (all the DLT series
>actually) expects an ANSI labeled tape, which is what ltf writes.
>Using a CompacTape III:
>
>ltf -cfB /dev/rmt#m 4096 tz88_v100_dgl.tz_image
>
>On the chance they are different, I'd be more inclined to use
>our V100 firmware instead of the Quantum version.


Jim Belonis _at_ washington.edu


>typically upgrading firmware just involves inserting the tape.
>the drive auto-detects that it is a firmware update and installs.

=======================================================================

Thanks to all mentioned above, and especially - to the angry subscriber.

V.S.G.


PS It would also be very helpful to know the secret place whence
the official tz88_v100_dgl.tz_image could be downloaded.
Received on Wed Apr 15 1998 - 17:29:20 NZST

This archive was generated by hypermail 2.4.0 : Wed Nov 08 2023 - 11:53:37 NZDT